Dosing control methods, devices, products and equipment in water treatment processes

Abstract

The present disclosure relates to the field of water treatment technology and provides a method, device, product, and equipment for controlling dosing in a water treatment process. The method comprises: obtaining characteristic values ​​of dosing prediction features in the water treatment process, the dosing prediction features including floc features, inlet parameter features, and outlet parameter features; inputting the characteristic values ​​of each dosing prediction feature into a dosing regression prediction model with the dosing prediction feature as a single input, to obtain a first dosing prediction value corresponding to each dosing prediction feature; inputting the characteristic values ​​of each dosing prediction feature into a dosing classification prediction model with the dosing prediction feature as a single input, to obtain a first dosing prediction level corresponding to each dosing prediction feature; and controlling the dosing in the water treatment process based on the first dosing prediction value and the first dosing prediction level corresponding to each dosing prediction feature. The present disclosure can improve the accuracy of dosing control.

Description

Technical Field

[0001] The present disclosure relates to the field of water treatment technology, and in particular to a dosing control method, a dosing control device, a computer program product, and an electronic device in a water treatment process. Background Art

[0002] Water treatment is the process of regulating raw water through physical, chemical, or biological means to achieve the required water quality standards for a specific use. For example, raw water can be treated with chemicals to remove impurities and make it suitable for drinking. Chemical addition is a crucial step in the water treatment process, determining the final effluent quality.

[0003] Related technologies can use machine learning models to process influent parameters, floc images, and other data to predict and control dosage. However, these technologies rely on a single model, and the output of this model cannot be verified, resulting in low prediction accuracy.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute prior art known to ordinary technicians in the field. Summary of the Invention

[0005] The purpose of the present disclosure is to provide a method, device, computer program product and electronic equipment for controlling dosing in a water treatment process, thereby improving the accuracy and reliability of dosing control in the water treatment process at least to a certain extent.

[0006] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by practice of the present disclosure.

[0007] According to a first aspect of the present disclosure, a method for controlling dosing in a water treatment process is provided, comprising: obtaining characteristic values ​​of dosing amount prediction features in the water treatment process, wherein the dosing amount prediction features include floc features, inlet parameter features, and outlet parameter features; inputting the characteristic values ​​of each dosing amount prediction feature into a preset dosing amount prediction model; inputting the characteristic values ​​of each dosing amount prediction feature into a dosing amount regression prediction model with the dosing amount prediction feature as a single input, and obtaining a first dosing amount prediction value corresponding to each dosing amount prediction feature according to the output of each dosing amount regression prediction model; inputting the characteristic values ​​of each dosing amount prediction feature into the preset dosing amount prediction model; In the dosing amount classification prediction model with the dosing amount prediction feature as a single input in the prediction model, the first dosing amount prediction level corresponding to each dosing amount prediction feature is obtained according to the output of each dosing amount classification prediction; the second dosing amount prediction value is determined based on the first dosing amount prediction value corresponding to each dosing amount prediction feature, and the second dosing amount prediction level is determined based on the first dosing amount prediction level corresponding to each dosing amount prediction feature; the target dosing amount prediction value is determined according to the overlapping relationship between the second dosing amount prediction value and the second dosing amount prediction level, and the dosing amount in the water treatment process is controlled based on the target dosing amount prediction value.

[0008] According to a second aspect of the present disclosure, a dosing control device for a water treatment process is provided, comprising: a characteristic value acquisition module configured to acquire characteristic values ​​of dosing amount prediction characteristics in the water treatment process, wherein the dosing amount prediction characteristics include floc characteristics, inlet parameter characteristics, and outlet parameter characteristics; a first prediction module configured to input the characteristic values ​​of each dosing amount prediction characteristic into a preset dosing amount prediction model and a dosing amount regression prediction model with the dosing amount prediction characteristic as a single input, and obtain a first dosing amount prediction value corresponding to each dosing amount prediction characteristic according to the output of each dosing amount regression prediction model; a second prediction module configured to input the characteristic values ​​of each dosing amount prediction characteristic into the In the preset dosing amount prediction model, in the dosing amount classification prediction model that uses the dosing amount prediction feature as a single input, the first dosing amount prediction level corresponding to each dosing amount prediction feature is obtained according to the output of each dosing amount classification prediction; the target prediction module is configured to determine the second dosing amount prediction value based on the first dosing amount prediction value corresponding to each dosing amount prediction feature, and determine the second dosing amount prediction level based on the first dosing amount prediction level corresponding to each dosing amount prediction feature; the dosing control module is configured to determine the target dosing amount prediction value based on the overlapping relationship between the second dosing amount prediction value and the second dosing amount prediction level, and control the dosing in the water treatment process based on the target dosing amount prediction value.

[0009] According to a third aspect of the present disclosure, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the steps of the method for controlling dosing in a water treatment process as described in the first aspect.

[0010] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the method for controlling dosing in a water treatment process as described in the first aspect of the above embodiment is implemented.

[0011] According to the fifth aspect of an embodiment of the present disclosure, an electronic device is provided, comprising: a processor; and a storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the dosing control method in the water treatment process as described in the first aspect of the above embodiment.

[0012] As can be seen from the above technical solutions, the dosing control method in a water treatment process, the dosing control device in a water treatment process, and the computer program product and electronic device for implementing the dosing control method in a water treatment process in the exemplary embodiments of the present disclosure have at least the following advantages and positive effects:

[0013] In the technical solutions provided by some embodiments of the present disclosure, the dosage value and dosage level are predicted respectively through multiple dosage prediction features, the final dosage prediction value is determined based on the overlapping relationship between the predicted dosage value and the dosage level, and the dosage in the water treatment process is controlled based on the dosage prediction value. Compared with the related art, on the one hand, the present disclosure predicts the dosage level through a classification prediction model, predicts the specific dosage value through a regression model, and determines the target dosage prediction value through the comprehensive result of the dosage level and the specific dosage value. In this way, the predicted dosage value can be verified based on the predicted dosage level, thereby improving the reliability of the prediction of the specific dosage value; on the other hand, the present disclosure predicts the dosage based on multiple dosage prediction features through a classification prediction model and a regression prediction model. The multiple dosage prediction features can adapt to the dosage prediction needs of different scenarios and improve the accuracy of the dosage prediction.

[0014] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification, are used to explain the principles of the present disclosure. Obviously, the drawings described below are only some embodiments of the present disclosure, and those skilled in the art can derive other drawings based on these drawings without inventive effort.

[0016] Figure 1 A schematic flow chart showing a method for controlling dosing in a water treatment process according to an exemplary embodiment of the present disclosure is provided;

[0017] Figure 2 A schematic diagram showing a water treatment system in an exemplary embodiment of the present disclosure is shown;

[0018] Figure 3 A schematic flow chart showing a method for determining a preset drug dosage prediction model in an exemplary embodiment of the present disclosure;

[0019] Figure 4 A schematic flow chart showing a method for determining different dosage levels in an exemplary embodiment of the present disclosure;

[0020] Figure 5 A flow chart showing a method for determining a target dosage prediction value based on an overlapping relationship in an exemplary embodiment of the present disclosure is shown;

[0021] Figure 6 A flow chart showing a method for controlling drug dosing according to a predicted target drug dosing amount in an exemplary embodiment of the present disclosure is shown;

[0022] Figure 7 A schematic diagram showing the composition of a dosing control device in a water treatment process according to an exemplary embodiment of the present disclosure;

[0023] Figure 8 A schematic structural diagram of an electronic device in an exemplary embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0024] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the present disclosure will be more comprehensive and complete and will fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or that other methods, components, devices, steps, etc. may be employed. In other cases, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0025] The terms "a", "an", "the" and "said" are used in this specification to indicate the presence of one or more elements / components / etc.; the terms "including" and "having" are used to express open-ended inclusion and mean that additional elements / components / etc. may exist in addition to the listed elements / components / etc.; the terms "first" and "second" etc. are used only as labels and are not intended to limit the quantity of their objects.

[0026] In addition, the accompanying drawings are merely schematic illustrations of the present disclosure and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0027] In the related art, although the dosage prediction in the water treatment process can be performed through machine learning models, the dosage prediction is performed through a single type of model. Even if the final prediction result is determined by multiple models, the prediction results cannot be mutually verified because the multiple models are of the same type, resulting in difficulty in ensuring the reliability and accuracy of the prediction results.

[0028] Based on this, the present disclosure provides a method for controlling dosing in a water treatment process to at least to some extent solve one or more problems in the above-mentioned related technologies.

[0029] For example, Figure 1 A schematic diagram showing a method for controlling the dosing of chemicals in a water treatment process according to an exemplary embodiment of the present disclosure is shown. Figure 1 , the method may include:

[0030] Step S110, obtaining characteristic values ​​of the dosing amount prediction characteristics in the water treatment process, wherein the dosing amount prediction characteristics include floc characteristics, influent parameter characteristics, and effluent parameter characteristics;

[0031] Step S120, inputting the characteristic value of each dosing amount prediction feature into a dosing amount regression prediction model in a preset dosing amount prediction model that uses the dosing amount prediction feature as a single input, and obtaining a first dosing amount prediction value corresponding to each dosing amount prediction feature based on the output of each dosing amount regression prediction model;

[0032] Step S130, inputting the characteristic value of each dosing amount prediction feature into the preset dosing amount prediction model, which uses the dosing amount prediction feature as a single input, and obtaining a first dosing amount prediction level corresponding to each dosing amount prediction feature based on the output of each dosing amount classification prediction;

[0033] Step S140, determining a second dosing amount prediction value based on the first dosing amount prediction value corresponding to each dosing amount prediction feature, and determining a second dosing amount prediction level based on the first dosing amount level corresponding to each dosing amount prediction feature;

[0034] Step S150 : determining a target dosing amount prediction value according to the overlapping relationship between the second dosing amount prediction value and the second dosing amount prediction level, and controlling the dosing amount in the water treatment process based on the target dosing amount prediction value.

[0035] exist Figure 1 In the technical solution provided by the illustrated embodiment, the dosage value and dosage level are predicted respectively through multiple dosage prediction features, the final dosage prediction value is determined based on the overlapping relationship between the predicted dosage value and the dosage level, and the dosage in the water treatment process is controlled based on the dosage prediction value. Compared with the related art, on the one hand, the dosage level is predicted by the classification prediction model, the specific dosage value is predicted by the regression model, and the target dosage prediction value is determined by the comprehensive result of the dosage level and the specific dosage value. In this way, the predicted dosage value can be verified based on the predicted dosage level, thereby improving the reliability of the prediction of the specific dosage value; on the other hand, the dosage prediction is performed based on the classification prediction model and the regression prediction model respectively based on multiple dosage prediction features. The multiple dosage prediction features can adapt to the dosage prediction needs of different scenarios and improve the accuracy of the dosage prediction.

[0036] Next, the specific implementation of "step S110, obtaining the characteristic value of the drug dosage prediction characteristic in the water treatment process" will be described in detail.

[0037] In order to more clearly illustrate the water treatment process of the present invention, first combine Figure 2 For illustration. For example, Figure 2 A schematic diagram of a water treatment system according to an exemplary embodiment of the present disclosure is shown. Figure 2 The water treatment system may include an inlet tank 21, a flocculation tank 22, a sedimentation tank 23, a filter tank 24 and an outlet tank 25. The inlet tank 21 can adjust the amount of incoming water and at the same time play a certain buffering role in the sudden changes in the quality of the raw water. The flocculation tank 22 is used to add flocculants, such as polyaluminum chloride, to the water to destabilize the fine suspended particles, colloids and other impurities in the water, and aggregate them to form larger flocs for subsequent sedimentation and separation. The sedimentation tank 23 uses gravity to cause the large particle flocs formed in the flocculation tank to settle to the bottom of the tank, thereby achieving solid-liquid separation and removing most of the suspended solids, mud and other impurities in the water. The filter tank 24 uses filter media, such as quartz sand, activated carbon, etc., to further remove impurities such as fine suspended particles, colloids, bacteria, etc. remaining in the water, thereby improving the clarity and transparency of the water and making the water quality reach higher standards. The outlet pool 25 is used to store treated clean water to provide a stable water supply guarantee for the water use link. At the same time, the water quality of the treated water can also be monitored in the outlet pool 25, such as testing pH, turbidity, suspended matter and other indicators to ensure that the outlet water quality meets the relevant standards and usage requirements.

[0038] For example, inlet water parameter detection equipment, such as inlet water pH detection equipment, inlet water temperature detection equipment, inlet water turbidity detection equipment, and inlet water flow detection equipment, can be installed between the inlet tank 21 and the flocculation tank 22. An underwater camera can be installed in the flocculation tank 22 to capture images of the flocs and obtain their characteristics. An underwater camera can also be installed in the sedimentation tank 23 to capture videos of floc sedimentation. A dosing pump can be installed between the inlet tank 21 and the flocculation tank 22 to add flocculant to the water.

[0039] In an exemplary embodiment, the dosage prediction features include floc features, inlet parameter features, and outlet parameter features. The dosage prediction features can be understood as features associated with the dosage and can be used to reflect or influence the dosage.

[0040] In an exemplary embodiment, floc characteristics can be determined using floc images captured by a camera. For example, an underwater camera can be installed in the flocculation tank, mounted on the inner wall of the tank, and can capture images of the flocs in the water. The camera can then send the captured floc images to a server. Upon receiving the floc images, the server performs feature extraction processing on the floc images to obtain floc characteristics. Floc characteristics may include floc area distribution, number of flocs per unit area, and average floc size.

[0041] For example, a floc recognition model can be pre-trained and floc features can be extracted using the floc recognition model. For example, a large number of historical floc images can be collected in advance and then manually annotated to identify the outline or boundary of each floc in the floc image, the floc area and floc particle size corresponding to the outline or boundary, and the number of flocs in the floc image as training data.

[0042] A machine learning model is trained using training data, enabling it to identify the outlines or boundaries of flocs in an image, as well as the area of ​​each floc, and to count the number of flocs in the image based on the identification results. Alternatively, an existing object detection model can be fine-tuned using the training data to train the object detection model to detect and identify flocs, thereby generating a floc recognition model. The floc recognition model can accurately identify flocs in an image containing flocs, determine the area and particle size of the identified flocs, and determine the number of flocs in the image. The recognition results of the floc recognition model are then analyzed and processed to obtain floc features. For example, the area distribution of the identified flocs (e.g., the percentage of each area), the average particle size of all flocs, and the number of flocs per unit area can be calculated based on the number of flocs in the image and the actual scene area indicated by the image. Floc features can also include other features, such as floc morphology and edge clarity, which are not specifically limited in this exemplary embodiment.

[0043] In an exemplary embodiment, the inlet water parameter characteristics include inlet water pH, inlet water flow rate, inlet water turbidity, and inlet water temperature. As previously mentioned, inlet water parameter detection devices such as inlet water pH detection devices, inlet water flow rate detection devices, inlet water turbidity detection devices, and inlet water temperature detection devices can be installed between the inlet water tank and the flocculation tank. These inlet water parameter detection devices can transmit the detected inlet water parameters to the server, and the server can obtain the inlet water parameter characteristics based on the data transmitted by these inlet water parameter detection devices.

[0044] In an exemplary embodiment, the effluent water parameter characteristics include effluent turbidity, effluent pH, and effluent suspended matter content. For example, effluent water parameter detection devices such as effluent turbidity detection devices, effluent pH detection devices, and effluent suspended matter content detection devices can be installed in the effluent water pool. These effluent water parameter detection devices can detect effluent water parameter characteristics and transmit the detected effluent water parameter characteristics to the server.

[0045] After obtaining the characteristic value of the dosing amount prediction feature, the server can determine the target dosing amount prediction value according to the subsequent steps S120 to S150, and then send a control instruction to the dosing pump according to the target dosing amount prediction value to control the amount of medicine added by the dosing pump.

[0046] In an exemplary embodiment, the water inlet parameters can be detected at regular intervals, and the most recently detected water inlet parameters can be compared with the currently detected water inlet parameters. When the change in any water inlet parameter exceeds the corresponding preset threshold, the camera is started to capture the floc image, and the current water inlet parameter characteristics, the captured floc image and the water outlet parameter characteristics are sent to the server, allowing the server to perform dosing prediction and control through the method disclosed in this disclosure.

[0047] In other words, in the present disclosure, it is possible to periodically collect inlet water parameters and then determine whether to re-predict the dosage based on the fluctuations of the inlet water parameters. For example, when the inlet water parameters change significantly, the dosage can be re-predicted and controlled. If the inlet water parameters do not change or the inlet water parameters do not change much from the previous inlet water parameters, the dosage corresponding to the previous inlet water parameters can be directly used for dosage control. This eliminates the need to re-predict the dosage at each sampling moment, improves dosage control efficiency, and also saves computing resources.

[0048] Of course, it is also possible to periodically collect inlet parameters, floc images, and outlet parameters, and then send the relevant data to a server to periodically perform predictive control of the dosage. That is, instead of determining fluctuations in the inlet parameters, predictive control of the dosage is performed repeatedly at a fixed time. This exemplary embodiment does not impose any particular limitation on this.

[0049] In an exemplary embodiment, the characteristic value of the dosing amount prediction feature can be understood as the specific value of each feature determined at the sampling moment, such as the inlet water temperature value, the inlet water pH value, etc.

[0050] Below, the specific implementation method of "step S120, inputting the characteristic value of each dosing dosage prediction feature into the preset dosing dosage prediction model and the dosing dosage regression prediction model with the dosing dosage prediction feature as the single input, and obtaining the first dosing dosage prediction value corresponding to each dosing dosage prediction feature according to the output of each dosing dosage regression prediction model" is described in detail.

[0051] For example, Figure 3 A flow chart showing a method for determining a preset dosage prediction model in an exemplary embodiment of the present disclosure is shown. Figure 3 The method may include steps S310 to S370.

[0052] In step S310 , a first label data set corresponding to each drug dosage prediction feature is generated based on the historical data corresponding to each drug dosage prediction feature and the historical drug dosage values ​​corresponding to the historical data.

[0053] For example, the characteristic values ​​of historical inlet water parameter features and the correct dosage values ​​for those characteristic values ​​can be collected to generate a first labeled dataset corresponding to the inlet water parameter features. Historical floc images can be collected, and floc features can be extracted from these images. Based on the extracted floc features and the correct dosage values ​​corresponding to the historical floc images, a first labeled dataset corresponding to the floc features can be generated. Historical outlet water parameter features and the correct dosage values ​​for those characteristic values ​​can be collected to generate a first labeled dataset corresponding to the outlet water parameter features. In other words, each dosage prediction feature corresponds to one labeled dataset, meaning the number of dosage prediction features corresponds to the number of first labeled datasets.

[0054] The correct dosage value can be understood as the dosage that can make the effluent water quality reach the preset standard under the characteristic value.

[0055] In an exemplary embodiment, historical data of drug dosage prediction features can be collected. If the drug dosage corresponding to the historical data is not a correct drug dosage value, it can be eliminated. That is, the collected historical data is filtered according to whether it corresponds to a normal drug dosage value, thereby obtaining a first label data set corresponding to each drug dosage prediction feature.

[0056] In step S320, a second label data set corresponding to each dosing dosage prediction feature is generated according to the historical data corresponding to each dosing dosage prediction feature and the dosing level to which the historical dosing dosage value corresponding to the historical data belongs in the different dosing levels corresponding to the dosing dosage prediction feature.

[0057] In an exemplary embodiment, different dosage levels corresponding to each dosage prediction feature may be predetermined, and different dosage levels correspond to different dosage ranges.

[0058] For example, different dosage prediction features correspond to the same number of dosage levels, and levels representing the same dosage degree have the same level identifier. For example, the dosage levels corresponding to each dosage prediction feature are divided into three levels: low, medium, and high. However, the dosage ranges indicated by the same dosage level corresponding to different dosage prediction features may be the same or different. For example, the dosage range indicated by the low-level dosage level corresponding to the influent parameter feature may be [a1, b1], while the dosage range indicated by the low-level dosage level corresponding to the floc feature may be [a2, b2]. Among them, a1, a2, b1, and b2 are all different values.

[0059] For example, the determination method of different dosage levels corresponding to any dosage prediction feature can refer to Figure 4 .like Figure 4As shown, the method may include steps S410 to S420. In which:

[0060] In step S410, historical data of the drug dosage prediction feature is collected, and the collected historical data is clustered.

[0061] For example, for each dosage prediction feature, historical data for that feature can be collected. Similarly, historical data can be filtered based on whether a correct dosage value is available to obtain historical data for that dosage prediction feature. For example, the inlet parameter feature corresponds to the first historical data, the floc feature corresponds to the second historical data, and the outlet parameter feature corresponds to the third historical data. Then, a first clustering is performed on the first historical data to obtain a first clustering result, a second clustering is performed on the second historical data to obtain a second clustering result, and a third clustering is performed on the third historical data to obtain a third clustering result.

[0062] The number of cluster categories for each dosing prediction feature is the same, i.e., the K value in the clustering algorithm is the same. Specifically, the K values ​​for the first cluster, the second cluster, and the third cluster are the same, such as 3. The specific K value can be determined based on experience or test results and is not particularly limited in this exemplary embodiment. For example, different K values ​​can be used, and the clustering results obtained with different K values ​​can be tested. The K value with the best test results can be selected as the final K value.

[0063] In step S420 , different drug dosage prediction levels are determined according to the drug dosage indicated by the historical data in each cluster category in the clustering result.

[0064] For example, the size relationship of the dosage levels between different dosage levels corresponding to different clustering categories can be determined based on the size relationship of the mean dosage indicated by the historical data in each clustering category in the clustering results; the dosage range indicated by the dosage level corresponding to the clustering category can be determined based on the minimum and maximum values ​​of the dosage indicated by the historical data in the clustering category.

[0065] Taking the K value of 3 as an example, each cluster category corresponds to a level, and there are 3 levels in total. The average value of the correct dosing values ​​corresponding to the historical data in cluster category 1 is c1, the average value of the correct dosing values ​​corresponding to the historical data in cluster category 2 is c2, and the average value of the correct dosing values ​​corresponding to the historical data in cluster category 3 is c3. The size relationship between c1, c2 and c3 is c1 less than c2 less than c3. Then cluster category 1 corresponds to the first dosing level, cluster category 2 corresponds to the second dosing level, and cluster category 3 corresponds to the third dosing level. The dosing degree of the first dosing level is less than the dosing degree of the second dosing level, and the dosing degree of the second dosing level is less than the third dosing level. For example, the first dosing level is a low dosing level, the second dosing level is a medium dosing level, and the third dosing level is a high dosing level.

[0066] For example, the dosage range indicated by each dosage level can be determined based on the minimum and maximum dosage values ​​indicated by the historical data in the cluster category corresponding to that dosage level. For example, if a dosage level has 100 historical data points, corresponding to 100 dosage values, the minimum and maximum values ​​of these 100 dosage values ​​constitute the dosage range indicated by that dosage level.

[0067] Taking the first, second, and third dosage levels described above as an example, the dosage ranges corresponding to the dosage levels determined based on the clustering results may overlap between the dosage ranges indicated by different dosage levels. In this case, manual verification and adjustment can be performed to ensure that the dosage ranges of adjacent dosage levels are continuous and non-overlapping. Alternatively, the dosage ranges corresponding to the dosage levels determined based on the clustering results can be used as the initial dosage ranges, and the initial dosage ranges can be adjusted according to the first preset rule to obtain the final dosage ranges corresponding to the dosage levels.

[0068] Exemplarily, the first preset rule may include: when there is an overlapping interval in the dosage range indicated by adjacent dosage levels, the overlapping interval is used as a buffer zone, that is, the dosage in the buffer zone belongs to the two adjacent dosage levels at the same time, and the dosage in the non-overlapping part belongs to the respective dosage levels. That is, the dosage range indicated by each dosage level is composed of the non-overlapping interval plus the buffer zone. In this way, when data is labeled, there are two labels for the dosage level of the historical data whose dosage value falls in the buffer zone. When making predictions later, the final output result is also selected by the category confidence. For example, the category with a confidence greater than 0.9 is the final predicted category. In this way, the problem of low prediction accuracy caused by the unreasonable division of the dosage level interval can be avoided to the greatest extent.

[0069] Exemplarily, the first preset rule may also include: dividing the overlapping interval equally and allocating it to the dosage range indicated by adjacent dosage levels, that is, dividing the overlapping interval into two parts with the median of the overlapping interval, and the final dosage range corresponding to the low level in the adjacent dosage levels is the minimum value in the corresponding initial dosage range to the median of the overlapping interval, and the final dosage range corresponding to the high level in the adjacent dosage levels is the median of the overlapping interval to the maximum value in the corresponding initial dosage range.

[0070] Similarly, if the initial dosage ranges corresponding to adjacent dosage levels are discontinuous, the second preset rule may be used for adjustment to obtain the final dosage range.

[0071] For example, the second preset rule may include: defining the range between the initial dosage ranges corresponding to adjacent dosage levels as a buffer zone, where the buffer zone also belongs to the adjacent dosage levels. Similarly, when labeling data, the dosage levels of historical data whose dosage values ​​fall within the buffer zone will have two labels.

[0072] Exemplarily, the second preset rule may also include: using the interval between the initial dosage ranges corresponding to adjacent dosage levels as a buffer zone, dividing the buffer zone into two by the median value of the buffer zone, and allocating the buffer zones to the adjacent dosage levels. That is, the final dosage range corresponding to the lower of the adjacent dosage levels is the range from the minimum value in the initial dosage range to the median value of the buffer zone, and the final dosage range corresponding to the higher of the adjacent dosage levels is the range from the median value of the buffer zone to the maximum value in the initial dosage range corresponding to the dosage level.

[0073] In an exemplary embodiment, if the dosage ranges indicated by non-adjacent dosage levels overlap, such as the overlapping dosage ranges corresponding to the low and high dosage levels described above, the K value is adjusted, clustering is performed again, and the dosage levels are re-determined based on the clustering results. Alternatively, the dosage ranges indicated by the non-adjacent dosage levels may be adjusted based on the dosage ranges indicated by the intermediate dosage levels between the non-adjacent dosage levels to ensure that the dosage ranges indicated by the non-adjacent dosage levels do not overlap. For example, the lowest value of the intermediate dosage level may be used as the upper limit of the lower dosage level among the non-adjacent dosage levels, and the maximum value of the intermediate dosage level may be used as the lower limit of the higher dosage level among the non-adjacent dosage levels.

[0074] Adjusting the initial dosing amount range based on the above method can ensure the rationality of the initial dosing amount range and ensure that the initial dosing amount range is continuous without any discontinuity. In this way, the target dosing amount prediction value can be determined in subsequent steps based on the overlapping relationship between the second dosing amount prediction value and the second dosing amount prediction level.

[0075] Through the above-mentioned steps S410 to S420, the dosage ranges indicated by different dosage levels corresponding to different dosage prediction features can be automatically determined based on historical data. Compared with the method of manually determining the dosage level, this clustering method of automatically determining the dosage level not only improves the efficiency of determining the dosage level, but also provides a basis for determining the dosage level, so that the determined dosage level has higher interpretability and reliability.

[0076] For example, after obtaining the dosage range indicated by different dosage levels corresponding to each dosage prediction feature, the first label data set corresponding to each dosage prediction feature can be copied to generate a first label copy data set. For the first label copy data set corresponding to any dosage prediction feature, the correct dosage value label in the first label copy data set is modified to the specific dosage level to which the correct dosage value belongs among the different dosage levels corresponding to the dosage prediction feature, thereby generating a second label data set.

[0077] Taking the water inlet parameter feature in the dosing prediction feature as an example, the first label data set A corresponding to the water inlet parameter feature is copied to generate the first label copy data set A1 corresponding to the water inlet parameter feature. According to the dosing range indicated by different dosing levels corresponding to the water inlet parameter feature, it is determined that the label of the training data in A1, that is, the specific correct dosing value, belongs to which range of the dosing range indicated by the different dosing levels corresponding to the water inlet parameter feature, thereby obtaining which level of the different dosing levels corresponding to the water inlet parameter feature the training data in A1 belongs to. The level is used as a new training label to generate a second label data set.

[0078] In other words, the training data in the first label data set and the second label data set can be the same, but their labels are different. The label of each training data in the first label data set is a specific dosage value, and the label of each training data in the second label data set is a dosage level.

[0079] In step S330 , a first number of candidate regression prediction models are trained based on the first label data set corresponding to each dosing amount prediction feature to obtain a second number of dosing regression prediction models.

[0080] In an exemplary embodiment, the first number is an integer greater than 1. The candidate regression prediction model may include any machine learning model capable of regression prediction, such as a neural network model, a regression model based on a support vector machine, etc., which is not particularly limited in this exemplary embodiment.

[0081] Taking the first number as 3 as an example, three candidate regression prediction models are trained based on the first labeled dataset corresponding to the influent parameter characteristics. Based on the training results, three dosing regression prediction models can be obtained. Three candidate regression prediction models are trained based on the first labeled dataset corresponding to the floc characteristics. Based on the training results, three dosing regression prediction models can be obtained. Based on the first labeled dataset corresponding to the effluent parameter characteristics, three candidate regression prediction models are trained based on the training results. Based on the training results, three dosing regression prediction models can be obtained. That is, a total of nine dosing regression prediction models can be obtained. That is, the second number and the first number are multiples, and the second number is three times the first number.

[0082] In step S340 , a third number of candidate classification prediction models are trained based on the second label data set corresponding to each dosing amount prediction feature to obtain a fourth number of dosing classification prediction models.

[0083] In an exemplary embodiment, the third number is an integer greater than 1. The candidate classification prediction model may include any machine learning model capable of classification prediction, such as a neural network classification model, a decision tree classification model, etc., which is not particularly limited in this exemplary embodiment.

[0084] Taking the third number of 4 as an example, similarly, based on each dosing amount prediction feature, 4 dosing classification prediction models can be obtained, and a total of 12 dosing classification prediction models can be obtained. That is, the fourth number and the third number are also in a multiple relationship, and the fourth number is 3 times the third number.

[0085] In step S350, a model test is performed on the second number of medication dosing regression prediction models, and a target medication dosing regression prediction model is determined from the second number of medication dosing regression prediction models according to the test results.

[0086] For example, a second number of medication dosing prediction models can be tested using a test data set, such as to determine their prediction accuracy. A medication dosing regression prediction model with a prediction accuracy greater than a preset threshold is then determined as a target medication dosing regression prediction model. The preset threshold can be determined based on demand or experience, and is not specifically limited in this exemplary embodiment.

[0087] Alternatively, for each dosing prediction feature corresponding to the dosing prediction model, the top N dosing prediction models ranked by test accuracy may be selected as candidate target dosing prediction models, and the target dosing prediction model may be obtained based on the set of candidate target dosing prediction models corresponding to each dosing prediction feature. N is greater than or equal to 1 and less than the first number.

[0088] In step S360, a model test is performed on the fourth number of medication classification prediction models, and a target medication classification prediction model is determined from the fourth number of medication classification prediction models according to the test result.

[0089] For example, the specific implementation of step S360 can refer to step S350 and will not be described again here.

[0090] In step S370, the preset drug dosage prediction model is determined according to the target drug dosage regression prediction model and the target drug dosage classification prediction model.

[0091] For example, a preset dosage prediction model can be obtained based on a combination of a target dosage regression prediction model and a target dosage classification prediction model. In other words, the preset dosage prediction model can be composed of multiple target dosage regression prediction models and multiple target dosage classification prediction models.

[0092] Through the above-mentioned steps S310 to S370, in the process of generating the preset dosage prediction model, data labeling and training are performed based on multiple dosage prediction features, thereby obtaining multiple target dosage regression prediction models and multiple target dosage classification prediction models. The multiple models can adapt to a variety of different water quality environments, thereby ensuring the accuracy and reliability of dosage prediction under different water quality environments.

[0093] For example, as mentioned above, each model in the preset drug dosage prediction model takes a single drug dosage prediction feature as input. Therefore, for each drug dosage prediction feature, the drug dosage prediction feature can be input into the drug dosage regression prediction model in the preset drug dosage prediction model with the drug dosage prediction feature as a single input, and the first drug dosage prediction value corresponding to the drug dosage prediction feature is obtained according to the output of the drug dosage regression prediction model.

[0094] The number of first dosage prediction values ​​that can be predicted by each dosage prediction feature is determined by the number of dosage regression prediction models in the preset dosage prediction model that use that dosage prediction feature as a single input. For example, if the inlet water parameter feature corresponds to two dosage regression prediction models, then two first dosage prediction values ​​can be predicted using the inlet water parameter feature. In other words, the number of first dosage prediction values ​​that can be predicted by each dosage prediction feature is the same as the number of target dosage regression prediction models corresponding to that dosage prediction feature in the preset dosage regression prediction model.

[0095] Below, the specific implementation method of "step S130, inputting the characteristic value of each dosing dosage prediction feature into the preset dosing dosage prediction model and the dosing dosage classification prediction model with the dosing dosage prediction feature as the single input, and obtaining the first dosing dosage prediction level corresponding to each dosing dosage prediction feature according to the output of each dosing dosage classification prediction" is described in detail.

[0096] Exemplarily, similarly, for each dosing dosage prediction feature, the dosing dosage prediction feature can be input into a preset dosing dosage prediction model and a dosing classification prediction model with the dosing dosage prediction feature as a single input, and the first dosing dosage prediction level corresponding to the dosing dosage prediction feature is obtained according to the output of the dosing classification prediction model.

[0097] For example, other specific implementations of step S130 may refer to the relevant content in step S120 and will not be described in detail here.

[0098] The specific implementation of "step S140, determining a second dosing amount prediction value based on the first dosing amount prediction value corresponding to each dosing amount prediction feature, and determining a second dosing amount prediction level based on the first dosing amount prediction level corresponding to each dosing amount prediction feature" is described in detail below.

[0099] Exemplarily, a specific implementation of step S140 may include: obtaining a second dosing dosage prediction value based on the average of the first dosing dosage prediction values ​​corresponding to each dosing dosage prediction feature, and obtaining a second dosing dosage prediction level based on the first dosing dosage level with the largest number among the first dosing dosage levels corresponding to each dosing dosage prediction feature.

[0100] For example, all the first dosing amount prediction values ​​obtained in step S120 may be averaged to obtain the second dosing amount prediction value, and the second dosing amount prediction level may be obtained based on the first dosing amount level that appears most frequently among the first dosing amount levels in step S130.

[0101] Next, the specific implementation of "step S150, determining a target dosing amount prediction value according to the overlapping relationship between the second dosing amount prediction value and the second dosing amount prediction level, and controlling the dosing amount in the water treatment process based on the target dosing amount prediction value" is described in detail.

[0102] Exemplarily, the overlapping relationship between the second dosage prediction value and the second dosage prediction level may include that the second dosage prediction value belongs to the dosage range indicated by the second dosage prediction level or that the second dosage prediction value does not belong to the dosage range indicated by the second dosage prediction level.

[0103] Taking the second dosage prediction level as the medium dosage level as an example, the dosage range indicated by the medium dosage level corresponding to each dosage prediction feature may be different. In an exemplary embodiment, as long as the second dosage prediction value falls within the dosage range indicated by the medium dosage level corresponding to any target dosage prediction feature, the second dosage prediction value is considered to fall within the dosage range indicated by the medium dosage level. If the second dosage prediction value does not fall within the dosage range indicated by the medium dosage level corresponding to any target dosage prediction feature, the second dosage prediction value is considered not to fall within the dosage range indicated by the medium dosage level. The target dosage prediction feature includes the dosage prediction feature corresponding to the first dosage prediction level and the dosage prediction feature having the same final second dosage prediction level. For example, there are 6 models in the target dosing classification prediction model. Among these 6 models, the prediction results of 4 models are medium dosing level, and the prediction results of 2 models are low dosing level, that is, the second predicted dosing level is medium dosing level. Then, the dosing prediction features input to the target dosing classification prediction model with 4 prediction results of medium dosing level are the target dosing prediction features.

[0104] Next, combine Figures 5 and 6 The specific implementation of step S150 is described in detail.

[0105] For example, Figure 5 A flow chart showing a method for determining a target dosage prediction value based on an overlapping relationship in an exemplary embodiment of the present disclosure is shown. Figure 5 , the method may include steps S510 to S540.

[0106] In step S510 , it is determined whether the second dosage prediction value is within the first dosage range indicated by the second dosage prediction level. If so, the process proceeds to step S520 ; otherwise, the process proceeds to step S530 .

[0107] In an exemplary embodiment, a database pre-stores dosage ranges corresponding to various dosage level indications corresponding to different dosage prediction features. For purposes of differentiation, each dosage range stored in the database is referred to as a first dosage range, and a dosage range obtained by adjusting each dosage range in the database is referred to as a second dosage range.

[0108] In step S520 , the second drug dosage prediction value is determined as the target drug dosage prediction value.

[0109] For example, as described above, as long as the second predicted dosage value falls within the first dosage range indicated by the second dosage prediction level corresponding to any target dosage prediction feature, the second predicted dosage value is considered to be within the first dosage range indicated by the second dosage prediction level. If the second predicted dosage value is within the first dosage range indicated by the second dosage prediction level, it indicates that the second predicted dosage value and the prediction result of the second dosage prediction level are consistent, that is, the dosage values ​​predicted by different models are consistent, indicating that the second predicted dosage value is reliable, and therefore the second predicted dosage value can be directly determined as the target predicted dosage value.

[0110] In step S530, the water flow rate and water turbidity detected at multiple detection points in the process from the raw water point to the water inlet point are obtained. When the water flow rate is greater than a first preset value and / or the water turbidity is greater than a second preset value, the first dosing amount range indicated by the second dosing amount prediction level is adjusted according to historical experience to obtain a second dosing amount range.

[0111] Taking the second dosage prediction level as the aforementioned medium dosage level as an example, if the second dosage prediction value does not fall within the first dosage range indicated by the medium dosage level corresponding to any target dosage prediction value, then the second dosage prediction value is considered to be outside the first dosage range indicated by the second dosage prediction level. Under normal circumstances, the prediction results of different models should be consistent. When the prediction results of different models are inconsistent, the prediction results of the models need to be re-evaluated.

[0112] For example, in the present disclosure, the dosage prediction value is constrained and verified by the dosage prediction level, so as to ensure the reliability and accuracy of the determined target dosage prediction value. Therefore, the division of dosage levels is crucial to the prediction results. When the prediction results of the classification prediction model and the regression prediction model are inconsistent, priority can be given to whether it is because the current water quality has undergone a sudden change, such as a sudden change in the turbidity of the influent, and the dosage ranges of each dosage level determined in advance may not cover the current water quality mutation condition, resulting in the problem of inconsistent prediction results of the classification prediction model and the regression prediction model. Therefore, in the present disclosure, when the second dosage prediction value does not belong to the first dosage range indicated by the second dosage prediction level of any target dosage prediction feature, the first dosage range indicated by the second dosage prediction level determined in advance can be dynamically adjusted to meet more complex emergencies as much as possible and improve the accuracy of dosage prediction.

[0113] In an exemplary embodiment, multiple monitoring points can be set up between the raw water point and the water inlet point to determine whether a sudden change in water quality has occurred due to sudden weather changes or other emergencies. The raw water point refers to the source of water to be treated in the water treatment system. For example, if the water treatment system treats water from a reservoir, the reservoir is the raw water point. The water inlet point refers to the entrance of the water treatment system's inlet tank.

[0114] For example, multiple detection points can be intermittently set between the raw water point and the water inlet point, each equipped with a turbidity meter and a flow meter. These meters can detect sudden changes in turbidity or flow. If the flow rate detected by the flow meter is greater than a first preset value and / or the flow rate detected by the turbidity meter is greater than a second preset value, a sudden change in water quality is determined, and the first dosage range indicated by the second dosage prediction level can be adjusted.

[0115] The first preset value and the second preset value can be determined according to needs and are not particularly limited in this exemplary embodiment. For example, the first preset value can be the maximum value of the inlet flow rate of all inlet parameter features in the training data multiplied by a coefficient greater than 1.

[0116] For example, the first dosage range indicated by the second dosage prediction level corresponding to the target dosage prediction feature can be adjusted based on historical experience. For example, if the target dosage prediction feature includes an inlet water parameter feature, a similarity calculation can be performed between the current inlet water parameter feature value and historical inlet water parameter feature values ​​stored in a database. Multiple historical inlet water parameter feature values ​​similar to the current inlet water parameter feature value can be selected, and the interval corresponding to the minimum and maximum historical dosage values ​​corresponding to the multiple historical inlet water parameter feature values ​​can be used as the second dosage range indicated by the second dosage prediction level corresponding to the inlet water parameter feature.

[0117] In step S540, a target drug dosage prediction value is determined based on the overlapping relationship between the second drug dosage range and the second drug dosage prediction value.

[0118] Exemplarily, an implementation of step S540 may include: when the second drug-dosing amount prediction value is within the second drug-dosing amount range, determining the second drug-dosing amount prediction value as the target drug-dosing amount prediction value; when the second drug-dosing amount prediction value is not within the second drug-dosing amount range, determining the target drug-dosing amount prediction value based on the minimum value in the first drug-dosing amount range indicated by the second drug-dosing amount prediction value and the second drug-dosing amount prediction level.

[0119] For example, when the second dosage prediction value is within the second dosage range indicated by the second dosage prediction level corresponding to any target dosage prediction feature, the second dosage prediction value can be determined as the target dosage prediction value. Otherwise, the minimum value between the first dosage range indicated by the second dosage prediction level corresponding to the target dosage prediction feature and the second dosage prediction value is determined as the target dosage prediction value. In this way, the dosage can be adjusted and controlled based on the target dosage prediction value first, and then adjusted based on the target dosage prediction value. Figure 6 The method shown determines whether further adjustment is required, thereby adjusting the dosage in a timely manner while avoiding the situation where excessive dosage causes secondary pollution to the water.

[0120] It should be noted that the above-mentioned adjustment of the first dosing amount range indicated by the second dosing amount prediction level to the second dosing amount range is only used in this comparison. The first dosing amount range is still stored in the database. The next time the dosing amount value is predicted, the first dosing amount range in the database is still used for determination.

[0121] In the present disclosure, dosing prediction can be performed using different types of models, and the prediction results of different types of models can verify and constrain each other, thereby improving the accuracy of the predicted dosing amount prediction value. At the same time, by using each dosing amount prediction feature as an input, the dosing amount can be predicted from multiple angles using multiple dosing amount prediction features, so that the model can adapt to a variety of different water quality environments, further improving the accuracy and reliability of the predicted dosing amount.

[0122] For example, Figure 6 A flow chart showing a method for controlling drug dosing according to a target drug dosing amount prediction value in an exemplary embodiment of the present disclosure is shown. Figure 6 The method may include steps S610 to S660.

[0123] In step S610 , the dosing pump is controlled to perform a dosing operation in the water treatment process according to a first target dosing amount indicated by the target dosing amount prediction value.

[0124] In an exemplary embodiment, the dosing pump can control the dosing amount by frequency. There is a mapping relationship between the frequency of the dosing pump and the dosing amount. The dosing pump frequency corresponding to the target dosing amount prediction value can be determined based on the mapping relationship, and the frequency can be sent to the dosing pump through an instruction to control the dosing pump to perform the dosing operation according to the frequency.

[0125] In step S620, in response to monitoring that the turbidity of water in the sedimentation tank decreases and the turbidity difference between adjacent sampling moments is less than a third preset value for the first time, the radar device and the video acquisition device acting on the sedimentation tank are simultaneously started.

[0126] For example, after controlling the dosing pump to perform dosing operations according to the predicted target dosing amount, the dosing effect can be detected by the floc state of the sedimentation tank, thereby monitoring whether the predicted target dosing amount is appropriate, and determining whether the dosing amount needs to be adjusted based on the monitoring situation.

[0127] In an exemplary embodiment, while sending a dosing control instruction to the dosing pump, a start control instruction may be sent to a turbidity detector in the sedimentation tank to start the turbidity detector in the sedimentation tank and obtain the turbidity in the sedimentation tank through the turbidity detector.

[0128] When the turbidity of water in the sedimentation tank is monitored by the turbidity meter and the turbidity difference between adjacent sampling moments is less than the third preset value for the first time under the downward trend, an opening control instruction is sent to the radar device and the video acquisition device acting on the sedimentation tank at the same time.

[0129] For example, when the water turbidity in the sedimentation tank shows a downward trend and tends to be stable for the first time, it means that the flocs in the sedimentation tank have begun to settle. At this time, the effect of dosing can be determined by the settling rate of the flocs.

[0130] For example, in the present disclosure, the settling velocity of flocs can be measured simultaneously using radar and video to ensure accurate determination of the settling velocity. Therefore, radar and video must measure the settling velocity of flocs within the same area during the same time period, so the radar and video acquisition devices can be activated simultaneously. At the same time, the installation locations of the radar and video acquisition devices must be arranged to ensure that their ranges of action overlap to the greatest extent possible, thereby ensuring that the radar and video acquisition devices measure the settling velocity of flocs in the same area, facilitating subsequent accurate fusion.

[0131] In step S630, a first average settling velocity of flocs in the sedimentation tank within a first preset time period is measured by a radar device.

[0132] In an exemplary embodiment, the first preset time period can be determined based on the radar activation time and the first preset duration, that is, the first preset time period is the period within the first preset duration after the radar is activated. Taking the first preset time period as 2 minutes as an example, the first preset time period is 2 minutes after the radar is activated.

[0133] For example, a radar device can transmit electromagnetic waves into a sedimentation tank. The waves are reflected by the flocs in the tank, and the radar receives the reflected waves. Due to the Doppler effect, the frequency of the reflected waves changes as the target object moves relative to the radar. If the target object moves toward the radar, the frequency of the reflected wave increases; if the target object moves away from the radar, the frequency of the reflected wave decreases. The change in frequency is proportional to the target object's speed. The radar measures the frequency difference between the transmitted and reflected waves, known as the Doppler shift, and combines this with the radar system parameters to determine the target object's speed.

[0134] For example, the radar signal can be filtered to avoid interference from water surface fluctuations. After filtering, the gain of the signals in different areas can be adjusted based on the intensity distribution of the radar signal, thereby facilitating subsequent analysis. Next, the time-domain radar signal can be converted into a frequency-domain signal, and the spectrum distribution of the signal can be obtained through Fourier transform. Extreme points are found in the spectrum, and different extreme points correspond to the reflection signals of floc groups with different speeds. The speed of each extreme point is determined based on the reflection frequency of each extreme point, and the speeds of all extreme points are weighted and averaged to obtain the average settling velocity of the flocs within the detection range during this measurement. The weight of the speed of each extreme point can be determined based on the energy of the extreme point. The greater the energy of the extreme point, the greater the weight of the corresponding speed. The speed of the different extreme points is weighted by the weight to obtain the average settling velocity.

[0135] In an exemplary embodiment, the average settling velocity of flocs within the detection range can be measured multiple times by radar signals within a first preset time period, and the average settling velocities measured multiple times within the first preset time period can be averaged again to obtain a first average settling velocity of the flocs in the sedimentation tank within the first preset time period.

[0136] In one exemplary embodiment, the radar device can also be used to measure the interface between the water in the sedimentation tank and the soil at the bottom of the sedimentation tank. The height of the interface can be used to determine whether to drain the sedimentation tank's sludge. For example, as flocs continue to settle, sludge gradually accumulates at the bottom of the sedimentation tank. The radar device can determine the interface between the sludge and water based on the different electromagnetic wave reflection characteristics of different media, thereby determining the height of the sludge interface. When the sludge interface reaches a preset height, an alarm can be triggered, prompting the sedimentation tank to drain, thereby preventing excessive sludge accumulation from affecting the sedimentation effect.

[0137] In other words, in the present disclosure, the radar device installed in the sedimentation tank can be used to measure the settling velocity of the flocs and can also be used to remind the sedimentation tank to discharge sewage. In this way, by reusing the radar device, the water treatment effect can be improved while reducing the hardware cost in the water treatment process.

[0138] In step S640, the video of the first preset time period in the sedimentation tank is collected by the video acquisition device, floc target tracking and video depth estimation are performed on the video, and the second average settling velocity of the flocs in the sedimentation tank during the first preset time period is determined based on the results of the floc target tracking and video depth estimation.

[0139] For example, a flocculent video can be sampled to obtain a flocculent image frame sequence. Then, in the first frame of the flocculent image frame sequence, all flocculents are detected using a target detection algorithm. Then, in the second and subsequent frames of the flocculent image frame sequence, the flocculents detected in the first frame are tracked using a multi-target tracking algorithm. The 2D pixel coordinates of the same flocculent in each flocculent image frame are determined based on the target tracking results. Simultaneously, video depth estimation can be performed on the flocculent image frame sequence. The depth value of the same flocculent in each flocculent image frame is determined based on the video depth estimation results. Using camera calibration parameters, the pixel coordinates and depth values ​​of the flocculent are converted to a 3D spatial position in the world coordinate system. The position component of the 3D spatial position in the flocculent settling direction is extracted. The flocculent settling displacement is determined based on the absolute value of the difference between this position component in the current frame and the reference frame of the current frame. The settling velocity between the current frame and the reference frame is then determined based on this settling displacement and the time interval between the current frame and the reference frame. The average settling velocity of the flocs in the first preset period is obtained according to the average settling velocity between each current frame and the reference frame, and the second average settling velocity of the flocs in the first preset period is obtained according to the average settling velocity of each successfully tracked floc in the first preset period.

[0140] The reference frame of the current frame may be a flocculation image frame that is located before the current frame and is N frames apart from the current frame, wherein N is determined based on experience or needs, such as N is 1, N is 2, or N is 0, etc. This exemplary embodiment does not impose any special limitation on this. N being 0 indicates that the reference frame of the current frame is the frame before the current frame.

[0141] In step S650, the first average sedimentation velocity and the second average sedimentation velocity are merged to obtain a target average sedimentation velocity.

[0142] In an exemplary embodiment, the first average sedimentation velocity and the second average sedimentation velocity can be fused based on a first weight corresponding to the first average sedimentation velocity and a second weight corresponding to the second average sedimentation velocity to obtain a target average sedimentation velocity. The first weight and the second weight can be determined based on needs and experience and are not particularly limited in this exemplary embodiment.

[0143] In step S660, the first target dosing amount is adjusted according to the difference between the target average sedimentation velocity and the preset sedimentation velocity to obtain a second target dosing amount, and the dosing pump is controlled to perform the dosing operation in the water treatment process according to the second target dosing amount.

[0144] In an exemplary embodiment, the preset settling velocity is determined based on the floc settling velocity corresponding to the target effluent water quality of the sedimentation tank. The target effluent water quality may include an effluent water quality that meets expected requirements, which is customized based on demand and is not specifically limited in this exemplary embodiment.

[0145] Exemplarily, the preset sedimentation velocity can be determined based on the correspondence between historical sedimentation velocity and effluent water quality. For example, the sedimentation velocity and the corresponding effluent water quality calculated each time can be recorded to obtain first recorded data. Then, the second recorded data whose effluent water quality meets the target effluent water quality can be selected from the first recorded data, and then the preset sedimentation velocity can be determined based on the sedimentation velocity in the second recorded data.

[0146] For example, the preset sedimentation velocity can be determined based on the average of the sedimentation velocities in the second recorded data, or the preset sedimentation velocity range can be determined based on the interval consisting of the minimum and maximum values ​​of the sedimentation velocities in the second recorded data. That is, the preset sedimentation velocity can also be a range interval. In the case where the preset sedimentation velocity is a range interval, if the target average sedimentation velocity falls within the range interval, it is considered that there is no difference between the target average sedimentation velocity and the preset sedimentation velocity, and the dosage does not need to be adjusted at this time. If the target average sedimentation velocity is less than the minimum value of the preset sedimentation velocity range and the absolute value of the difference with the minimum value is greater than the fourth preset value, the first adjustment strategy is executed; if the target average sedimentation velocity is greater than the maximum value of the preset sedimentation velocity range and the absolute value of the difference with the maximum value is greater than the fourth preset value, the second adjustment strategy is executed. When the preset sedimentation velocity is a specific value, if the target average sedimentation velocity is less than the preset sedimentation velocity and the absolute value of the difference with the preset sedimentation velocity is greater than the fourth preset value, the first adjustment strategy is executed; if the target average sedimentation velocity is greater than the preset sedimentation velocity and the absolute value of the difference with the preset sedimentation velocity is greater than the fourth preset value, the second adjustment strategy is executed.

[0147] In an exemplary embodiment, the first adjustment strategy includes: increasing the current dosing amount of the dosing pump according to a preset increase proportional coefficient. For example, the preset increase proportional coefficient can be greater than 1, such as 1.05. In this way, the value obtained by multiplying the current dosing amount by 1.05 is determined as the new dosing amount each time. After re-dosing, the target average sedimentation velocity is re-determined according to the above method, and then it is determined whether it needs to be adjusted again. This cycle is repeated until the difference between the newly determined target average sedimentation velocity and the preset sedimentation velocity meets the preset conditions, and the adjustment is stopped.

[0148] In an exemplary embodiment, the second adjustment strategy includes: reducing the current dosing amount of the dosing pump according to a preset reduction proportional coefficient. For example, the preset reduction proportional coefficient can be less than 1, such as 0.95. In this way, the value obtained by multiplying the current dosing amount by 0.95 is determined as the new dosing amount each time. After re-dosing, the target average sedimentation velocity is re-determined according to the above method, and then it is determined whether it needs to be adjusted again. This cycle is repeated until the difference between the newly determined target average sedimentation velocity and the preset sedimentation velocity meets the preset conditions, and the adjustment is stopped.

[0149] In other words, if the flocs settle faster after adding the drug, it means that the amount of drug added is large, causing the flocs to further aggregate to form larger flocs, and the settling rate may be accelerated. If the flocs settle slower, it means that the amount of drug added is small.

[0150] The preset condition includes that the absolute value of the difference between the target average sedimentation velocity and the preset sedimentation velocity is less than a fourth preset value.

[0151] In an exemplary embodiment, the dosage when adjustment is stopped can be associated with the inlet parameter characteristics, floc characteristics and outlet water quality parameter characteristics to generate an updated database, and each model in the above-mentioned preset dosage prediction model can be regularly updated based on the updated database.

[0152] Through the above steps S610 to S660, the floc settling velocity can be determined in a variety of ways, improving the accuracy of the floc settling velocity determination. Based on the floc settling velocity, the dosing treatment effect of the target dosage prediction value can be judged, and the dosage can be adjusted based on the judgment result to more accurately control the dosage. At the same time, adjusting the dosage based on the floc condition during the settling stage can improve the timeliness of the dosage adjustment.

[0153] In the present disclosure, by using multiple dosage prediction features as single inputs to predict dosage, the dosage can be predicted and verified from multiple angles, so that the dosage can be predicted in different water quality scenarios, thereby improving the accuracy of the dosage prediction. At the same time, the dosage prediction is performed simultaneously through the dosage classification prediction model and the dosage regression prediction model, and the prediction results of the dosage regression prediction model are verified and constrained based on the output results of the dosage classification prediction model. Not only can the specific dosage value be predicted, but the accuracy and reliability of the predicted specific dosage value can also be improved, thereby assisting in improving the stability and reliability of the dosage control in the water treatment process. In addition, since the accuracy of the dosage control is improved by the method disclosed in the present disclosure, the degree of manual participation in the dosage process can be reduced, and the automation of the water treatment process can be achieved to a higher degree.

[0154] Furthermore, it should be noted that the aforementioned figures are merely illustrative of the processes included in the method according to exemplary embodiments of the present invention and are not intended to be limiting. It is readily understood that the processes illustrated in the aforementioned figures do not indicate or limit the temporal order of these processes. Furthermore, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0155] Furthermore, the exemplary embodiment of the present disclosure also provides a dosing control device in a water treatment process. Figure 7As shown, the dosing control device 700 in the water treatment process includes the following program modules: a characteristic value acquisition module 710, which is configured to obtain the characteristic value of the dosing amount prediction feature in the water treatment process, and the dosing amount prediction feature includes a floc feature, an inlet water parameter feature and an outlet water parameter feature; a first prediction module 720, which is configured to input the characteristic value of each dosing amount prediction feature into a preset dosing amount prediction model and a dosing amount regression prediction model with the dosing amount prediction feature as a single input, and obtain the first dosing amount prediction value corresponding to each dosing amount prediction feature according to the output of each dosing amount regression prediction model; a second prediction module 730, which is configured to input the characteristic value of each dosing amount prediction feature into the preset dosing amount prediction model and the dosing amount regression prediction model with the dosing amount prediction feature as a single input, and obtain the first dosing amount prediction value corresponding to each dosing amount prediction feature according to the output of each dosing amount regression prediction model; Assume that in a dosing dosage prediction model, in which the dosing dosage prediction feature is used as a single input, the first dosing dosage prediction level corresponding to each dosing dosage prediction feature is obtained according to the output of each dosing dosage classification prediction; the target prediction module 740 is configured to determine the second dosing dosage prediction value based on the first dosing dosage prediction value corresponding to each dosing dosage prediction feature, and to determine the second dosing dosage prediction level based on the first dosing dosage level corresponding to each dosing dosage prediction feature; the dosing control module 750 is configured to determine the target dosing dosage prediction value according to the overlapping relationship between the second dosing dosage prediction value and the second dosing dosage prediction level, and to control the dosage in the water treatment process based on the target dosing dosage prediction value.

[0156] In an exemplary embodiment, the target dosing amount prediction value is determined based on the overlapping relationship between the second dosing amount prediction value and the second dosing amount prediction level, including: when the second dosing amount prediction value is within the first dosing amount range indicated by the second dosing amount prediction level, determining the second dosing amount prediction value as the target dosing amount prediction value; when the second dosing amount prediction value is not within the first dosing amount range indicated by the second dosing amount prediction level, obtaining the water flow and water turbidity detected at multiple detection points in the process from the raw water point to the water inlet point, when the water flow is greater than the first preset value and / or the water turbidity is greater than the second preset value, adjusting the first dosing amount range indicated by the second dosing amount prediction level according to historical experience to obtain the second dosing amount range, and determining the target dosing amount prediction value according to the overlapping relationship between the second dosing amount range and the second dosing amount prediction value.

[0157] In an exemplary embodiment, determining the target dosing amount prediction value based on the overlapping relationship between the second dosing amount range and the second dosing amount prediction value includes: when the second dosing amount prediction value is within the second dosing amount range, determining the second dosing amount prediction value as the target dosing amount prediction value; when the second dosing amount prediction value is not within the second dosing amount range, determining the target dosing amount prediction value based on the minimum value in the first dosing amount range indicated by the second dosing amount prediction value and the second dosing amount prediction level.

[0158] In an exemplary embodiment, the method for determining the preset drug dosage prediction model includes: generating a first label data set corresponding to each drug dosage prediction feature based on historical data corresponding to each drug dosage prediction feature and historical drug dosage values ​​corresponding to the historical data;

[0159] A second label data set corresponding to each dosing dosage prediction feature is generated based on the historical data corresponding to each dosing dosage prediction feature and the dosing dosage level to which the historical dosing dosage value corresponding to the historical data belongs in the different dosing dosage levels corresponding to the dosing dosage prediction feature; a first number of candidate regression prediction models are trained based on the first label data set corresponding to each dosing dosage prediction feature to obtain a second number of dosing regression prediction models, and the second number is a multiple of the first number; a third number of candidate classification prediction models are trained based on the second label data set corresponding to each dosing dosage prediction feature to obtain a fourth number of dosing classification prediction models, and the fourth number is a multiple of the third number; a model test is performed on the second number of dosing regression prediction models, and a target dosing regression prediction model is determined from the second number of dosing regression prediction models according to the test results; a model test is performed on the fourth number of dosing classification prediction models, and a target dosing classification prediction model is determined from the fourth number of dosing classification prediction models according to the test results; and the preset dosing dosage prediction model is determined based on the target dosing regression prediction model and the target dosing classification prediction model.

[0160] In an exemplary embodiment, a method for determining different dosing levels corresponding to any dosing prediction feature includes: collecting historical data of the dosing prediction feature, clustering the collected historical data; and determining different dosing prediction levels based on the dosing indicated by the historical data in each cluster category in the clustering results.

[0161] In an exemplary embodiment, the method of determining different dosing amount prediction levels based on the dosing amount indicated by historical data in each cluster category in the clustering results includes: determining the size relationship of the dosing amount between different dosing amount levels corresponding to different cluster categories based on the size relationship of the mean value of the dosing amount indicated by historical data in each cluster category in the clustering results; and determining the dosing amount range indicated by the dosing amount level corresponding to the cluster category based on the minimum and maximum values ​​of the dosing amount indicated by the historical data in the cluster category.

[0162] In an exemplary embodiment, the control of the dosage of the water treatment process based on the target dosage prediction value includes: controlling the dosage pump to perform the dosage operation in the water treatment process according to the first target dosage indicated by the target dosage prediction value; in response to monitoring that the turbidity of water in the sedimentation tank decreases and the turbidity difference between adjacent sampling moments is less than a third preset value for the first time, simultaneously starting a radar device and a video acquisition device acting on the sedimentation tank; measuring a first average settling velocity of flocs in the sedimentation tank within a first preset time period by the radar device; collecting a video of the sedimentation tank within the first preset time period by the video acquisition device, and analyzing the video Frequent floc target tracking and video depth estimation are performed, and a second average settling velocity of the flocs in the sedimentation tank within the first preset time period is determined based on the results of the floc target tracking and video depth estimation; the first average settling velocity and the second average settling velocity are merged to obtain a target average settling velocity; according to the difference between the target average settling velocity and the preset settling velocity, the first target dosing amount is adjusted to obtain a second target dosing amount, and the dosing pump is controlled to perform a dosing operation in the water treatment process according to the second dosing amount; wherein the preset settling velocity is determined based on the floc settling velocity corresponding to the target effluent water quality of the sedimentation tank.

[0163] The specific details of each part of the above-mentioned device have been described in detail in the implementation method part. The undisclosed details can be found in the implementation method part, so they will not be repeated here.

[0164] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the exemplary embodiments of the present disclosure, the features and functions of two or more modules or units described above can be concretized in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0165] Furthermore, although the steps of the method of the present disclosure are described in a particular order in the accompanying drawings, this does not require or imply that the steps must be performed in this particular order, or that all steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0166] The exemplary embodiments of the present disclosure further provide a computer program product, which includes a computer program, and when the computer program is executed by a processor, implements the above-mentioned method for controlling dosing in a water treatment process.

[0167] In one embodiment, a computer program product may be a tangible product containing a computer program, such as a computer-readable storage medium storing the computer program. The computer-readable storage medium may be a storage medium based on electrical, magnetic, optical, electromagnetic, infrared, or other signals, including but not limited to random access memory (RAM), read-only memory (ROM), magnetic tape, floppy disk, flash memory (Flash), hard disk drive (HDD), solid-state drive (SSD), and the like. Exemplarily, the computer program product may be implemented as a non-volatile storage medium storing the computer program, such as a read-only memory (ROM) or NAND flash memory.

[0168] In one embodiment, the computer program product may be an intangible product containing a computer program. For example, the computer program product may be implemented as a virtual digital product, such as a digital file such as an executable file or installation package storing the computer program.

[0169] The code of a computer program can be written in one or more programming languages, such as C, Java, C++, Python, and the like. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device can be connected to the user's computing device via any type of network, such as a local area network (LAN) or wide area network (WAN), or can be connected to an external computing device (e.g., via an internet connection provided by a carrier).

[0170] Computer programs can be carried or transmitted via electrical, magnetic, optical, electromagnetic, infrared, or other signals. Electronic devices can convert signals carrying computer programs into digital signals to execute the computer programs. When the computer program is executed on an electronic device, its code causes the electronic device (more specifically, the processor of the electronic device) to execute the method steps of various exemplary embodiments of the present disclosure.

[0171] The exemplary embodiments of the present disclosure also provide an electronic device. The electronic device may include a processor and a memory. The memory stores executable instructions for the processor, such as a computer program. The processor executes the executable instructions to perform the method steps of various exemplary embodiments of the present disclosure. The electronic device may also include a display for displaying a graphical user interface.

[0172] Reference below Figure 8 , the electronic device is exemplarily described in the form of a general-purpose computing device. It should be understood that Figure 8 The electronic device 800 shown is only an example and should not limit the functions and scope of use of the embodiments of the present disclosure.

[0173] like Figure 8 As shown, the electronic device 800 may include a processor 810 , a memory 820 , a bus 830 , an I / O (input / output) interface 840 , a network adapter 850 , and a display 860 .

[0174] The memory 820 may include volatile memory, such as RAM 821 and cache unit 822, and may also include non-volatile memory, such as ROM 823. The memory 820 may also include one or more program modules 824. Such program modules 824 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. For example, the program modules 824 may include the modules in the aforementioned devices.

[0175] The processor 810 may include one or more processing units. For example, the processor 810 may include an AP (Application Processor), a modem processor, a GPU (Graphics Processing Unit), an ISP (Image Signal Processor), a controller, an encoder, a decoder, a DSP (Digital Signal Processor), a baseband processor and / or an NPU (Neural-Network Processing Unit), etc.

[0176] The processor 810 may be configured to execute executable instructions stored in the memory 820 , such as executing the aforementioned method for controlling dosing in a water treatment process.

[0177] The bus 830 is used to realize the connection between different components of the electronic device 800 and may include a data bus, an address bus, and a control bus.

[0178] The electronic device 800 can communicate with one or more external devices 900 (eg, a keyboard, a mouse, an external controller, etc.) through the I / O interface 840 .

[0179] The electronic device 800 can communicate with one or more networks via the network adapter 850. For example, the network adapter 850 can provide mobile communication solutions such as 3G / 4G / 5G, or wireless communication solutions such as wireless LAN, Bluetooth, and near-field communication. The network adapter 850 can communicate with other modules of the electronic device 800 via the bus 830.

[0180] The electronic device 800 may display a graphical user interface via the display 860 , such as an interface for displaying a drug dosage prediction value.

[0181] although Figure 8 Not shown, other hardware and / or software modules may also be provided in the electronic device 800, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0182] Furthermore, the figures above are merely illustrative of the processes included in the methods according to exemplary embodiments of the present disclosure and are not intended to be limiting. It is readily understood that the processes illustrated in the figures above do not indicate or limit the temporal order of these processes. Furthermore, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0183] As can be seen from the above, the technical solutions of the present disclosure can be implemented as methods, devices, systems, computer program products, storage media, electronic devices, etc. Those skilled in the art will appreciate that various aspects of the present disclosure can be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation that combines hardware and software aspects, such as "circuits," "modules," or "systems," respectively.

[0184] It should be understood that the present disclosure is not limited to the specific method steps or structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from the scope thereof. Those skilled in the art will easily think of other embodiments based on the specific embodiments provided by the present disclosure. Therefore, the specific embodiments provided by the present disclosure are merely exemplary, and the scope and spirit of the present disclosure are indicated by the claims, which should cover any variations, uses or adaptive changes of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary technical means in the field of the present technology that are not disclosed in the present disclosure.

Claims

1. A method for controlling dosing in a water treatment process, characterized in that: include: Obtaining characteristic values ​​of dosing amount prediction features in a water treatment process, wherein the dosing amount prediction features include floc characteristics, influent parameter characteristics, and effluent parameter characteristics; Inputting the characteristic value of each dosage prediction feature into a preset dosage prediction model, which uses the dosage prediction feature as a single input, and obtaining a first dosage prediction value corresponding to each dosage prediction feature according to the output of each dosage regression prediction model; Inputting the characteristic value of each dosing amount prediction feature into the preset dosing amount prediction model and the dosing amount classification prediction model with the dosing amount prediction feature as a single input, and obtaining the first dosing amount prediction level corresponding to each dosing amount prediction feature according to the output of each dosing amount classification prediction; Determine a second dosage prediction value based on the first dosage prediction value corresponding to each dosage prediction feature, and determine a second dosage prediction level based on the first dosage level corresponding to each dosage prediction feature; A target dosing amount prediction value is determined according to the overlapping relationship between the second dosing amount prediction value and the second dosing amount prediction level, and the dosing amount in the water treatment process is controlled based on the target dosing amount prediction value.

2. The method according to claim 1, characterized in that The step of determining a target drug dosage prediction value based on an overlapping relationship between the second drug dosage prediction value and the second drug dosage prediction level includes: When the second drug dosage prediction value is within a first drug dosage range indicated by the second drug dosage prediction level, determining the second drug dosage prediction value as the target drug dosage prediction value; When the second dosing amount prediction value is not within the first dosing amount range indicated by the second dosing amount prediction level, the water flow rate and water turbidity detected at multiple detection points in the process from the raw water point to the water inlet point are obtained. When the water flow rate is greater than the first preset value and / or the water turbidity is greater than the second preset value, the first dosing amount range indicated by the second dosing amount prediction level is adjusted according to historical experience to obtain the second dosing amount range. According to the overlapping relationship between the second dosing amount range and the second dosing amount prediction value, the target dosing amount prediction value is determined.

3. The method according to claim 2, characterized in that The determining of the target dosage prediction value according to the overlapping relationship between the second dosage range and the second dosage prediction value includes: When the second drug dosage prediction value is within the second drug dosage range, determining the second drug dosage prediction value as the target drug dosage prediction value; When the second predicted dosage value is not within the second dosage range, a target predicted dosage value is determined according to the second predicted dosage value and a minimum value in the first dosage range indicated by the second predicted dosage level.

4. The method according to claim 1, wherein The method for determining the preset dosage prediction model includes: Generate a first label data set corresponding to each drug dosage prediction feature based on historical data corresponding to each drug dosage prediction feature and historical drug dosage values ​​corresponding to the historical data; Generate a second label data set corresponding to each drug dosage prediction feature according to historical data corresponding to each drug dosage prediction feature and the drug dosage level to which the historical drug dosage value corresponding to the historical data belongs among the different drug dosage levels corresponding to the drug dosage prediction feature; Training a first number of candidate regression prediction models based on the first label data set corresponding to each dosing dosage prediction feature to obtain a second number of dosing dosage regression prediction models, where the second number is a multiple of the first number; Training a third number of candidate classification prediction models based on the second label data set corresponding to each dosing dosage prediction feature to obtain a fourth number of dosing classification prediction models, where the fourth number is a multiple of the third number; performing a model test on the second number of medication dosing regression prediction models, and determining a target medication dosing regression prediction model from the second number of medication dosing regression prediction models according to the test results; performing a model test on the fourth number of medication classification prediction models, and determining a target medication classification prediction model from the fourth number of medication classification prediction models according to the test results; The preset drug dosage prediction model is determined based on the target drug dosage regression prediction model and the target drug dosage classification prediction model.

5. The method according to claim 4, characterized in that The methods for determining different dosage levels corresponding to any dosage prediction feature include: Collecting historical data of the dosage prediction feature and clustering the collected historical data; Different dosage amount prediction levels are determined according to the dosage amount indicated by the historical data in each cluster category in the clustering result.

6. The method according to claim 5, characterized in that Determining different dosage prediction levels according to the dosage indicated by the historical data in each cluster category in the clustering result includes: Determine the magnitude relationship between different dosage levels corresponding to different cluster categories based on the magnitude relationship between the mean values ​​of dosages indicated by the historical data in each cluster category in the clustering result; The dosage range indicated by the dosage level corresponding to the cluster category is determined according to the minimum value and the maximum value of the dosage indicated by the historical data in the cluster category.

7. The method according to any one of claims 1 to 6, characterized in that The controlling of the dosage of the water treatment process based on the target dosage prediction value includes: controlling the dosing pump to perform a dosing operation in the water treatment process according to a first target dosing amount indicated by the target dosing amount prediction value; In response to monitoring that the turbidity of water in the sedimentation tank decreases and the turbidity difference between adjacent sampling moments is less than a third preset value for the first time, simultaneously activating a radar device and a video acquisition device acting on the sedimentation tank; measuring a first average settling velocity of flocs in the sedimentation tank within a first preset time period by the radar device; collecting, by the video acquisition device, a video of the sedimentation tank within the first preset time period, performing floc target tracking and video depth estimation on the video, and determining a second average settling velocity of the flocs in the sedimentation tank within the first preset time period based on the results of the floc target tracking and video depth estimation; fusing the first average sedimentation velocity and the second average sedimentation velocity to obtain a target average sedimentation velocity; adjusting the first dosing amount according to the difference between the target average settling velocity and the preset settling velocity to obtain a second dosing amount, and controlling the dosing pump to perform a dosing operation in the water treatment process according to the second dosing amount; The preset settling velocity is determined according to the floc settling velocity corresponding to the target effluent water quality of the sedimentation tank.

8. A dosing control device in a water treatment process, characterized in that: include: a characteristic value acquisition module configured to acquire characteristic values ​​of dosage prediction characteristics in a water treatment process, wherein the dosage prediction characteristics include floc characteristics, influent parameter characteristics, and effluent parameter characteristics; The first prediction module is configured to input the characteristic value of each dosage prediction feature into a dosage regression prediction model in a preset dosage prediction model, which uses the dosage prediction feature as a single input, and obtain a first dosage prediction value corresponding to each dosage prediction feature according to the output of each dosage regression prediction model; The second prediction module is configured to input the characteristic value of each dosing amount prediction feature into the preset dosing amount prediction model, which uses the dosing amount prediction feature as a single input, and obtain a first dosing amount prediction level corresponding to each dosing amount prediction feature according to the output of each dosing amount classification prediction; a target prediction module configured to determine a second drug dosage prediction value based on the first drug dosage prediction value corresponding to each drug dosage prediction feature, and to determine a second drug dosage prediction level based on the first drug dosage level corresponding to each drug dosage prediction feature; The dosing control module is configured to determine a target dosing amount prediction value according to the overlapping relationship between the second dosing amount prediction value and the second dosing amount prediction level, and control the dosing amount in the water treatment process based on the target dosing amount prediction value.

9. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, causes the one or more processors to implement the method according to any one of claims 1 to 7.

Images